The present invention relates to packet switching systems and network configurations, and in particular to an apparatus and a network configuration suitable for signal processing in asynchronous transfer mode (ATM) switching networks for switching fixed length packets.
Currently, in ITU-T (International Telecommunication Union--Telecommunication standardization sector), standardization works of various recommendations for implementing the Broadband ISDN are under way. The Broadband ISDN aims at providing service with a line rate of 156 Mbits/sec, for example, to users by adopting an ATM switching method.
As for a method for implementing signal processing in switching systems implementing the ATM switching method (hereafter referred to as ATM switching systems), a method described in U.S. Pat. No. 5,214,642 is known. According to this method, signal processing is implemented by connecting subscriber lines and trunks to a signal processor via an ATM switch. Since the signal processor is shared by a plurality of subscriber lines and trunks via the ATM switch, an economical ATM switching system is implemented.
As for the ATM switching system for implementing the Broadband ISDN, signals of a large number of users undergo high speed processing and consequently a system of high reliability is needed. As a measure for maintaining communication in the event of a failure of a part of the ATM switching system, therefore, reliability improvement such as a conventionally known redundant configuration in the switching system is attempted as represented by "duplication," for example. In the "duplicated" configuration, functional blocks forming the ATM switching system, such as the switch, signal processor and processor, are duplicated block by block. In the case where one of the duplicated blocks fails, switchover to the other is conducted. Even when a fault has occurred, therefore, the normal state is quickly restored.
Furthermore, in the conventional ATM switching system, a signal processor is installed for each ATM switching system and a function of setting/releasing an information channel in response to a request from a user is implemented. The SVC (Signalling Virtual Channel) function is thus implemented.
In the conventional known ATM switching network, contrivances have been effected. For example, the signal processor is shared in the ATM switching system. Configuring the ATM switching network with due regard to securement of reliability and provision of a wide variety of communication services causes complicated configurations of switching systems and the whole network because of adoption of redundant configurations, resulting in an increased cost. As for the process of spread of the Broadband ISDN, large-scale switching networks forming a trunk line network having a large capacity are introduced at the beginning in the same way as the process of construction of the conventional networks. In this case, however, the number of subscribers subject to signal processing is very large. Even if there is a cost increase as described above, therefore, the cost increase per subscriber is suppressed to a relatively small value and consequently no problems are posed. When introduction of the Broadband ISDN has finally advanced to introduction thereof to small-sized communication facilities and communication networks, such as LANs and private switching networks, accommodating a small number of subscribers, a cost increase per subscriber becomes very large provided that a system configuration similar to that of a large scale switching network forming a trunk network as described above is used. Because communication services and traffic conditions applied to these small scale switching networks are different and diverse, and a smaller number of subscribers are accommodated by these small scale switching networks.
Owing to the progress of hardware techniques, there is presently little probability that hardware such as switches gets out of order. On the other hand, the number of steps of software for executing call control and the like has become large in order to implement various functions attendant upon introduction of intelligent networks (INs) for providing various communication services. The proportion of faults caused by software is thus increasing. In such a situation, duplication of each device in the switching network is not effective in many cases for relief in a fault caused by software and improvement of reliability and it increases the device complexity. For forming and spreading broadband communication systems and communication networks hereafter, it is important to take a reliability improving measure narrowed down to software faults.
Furthermore, in order to cope with the increase of IN functions, the processor executing the call control and the like is needed to have a greater processing capability year after year. In the present configuration of switching systems and switching networks, however, it is not easy to replace the processor of a switching system with that having a higher performance and a cost for replacement is also needed.